16:00:02 #startmeeting integration 16:00:02 Meeting started Thu Jun 15 16:00:02 2017 UTC. The chair is jamoluhrsen. Information about MeetBot at http://ci.openstack.org/meetbot.html. 16:00:02 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:02 The meeting name has been set to 'integration' 16:04:27 #topic release 16:04:51 #info boron sr4 will be delayed by a day or two because of a blocker in netvirt 16:04:58 #info netvirt fix coming tomorrow 16:05:47 #info carbon SR1 in 3 weeks 16:06:16 #info 7/2 is the cut-off for carbon sr1 16:06:46 #info nitrogen will have a painful next two weeks. 16:07:12 #info all projects are moving to odlparent 1.9.0 16:07:54 #info infrautils is back now 16:08:17 #info next week we want to move projects to odlparent 2.0.0 16:09:02 #info in summary we have two weeks of trouble in nitrogen while we have two weeks to get carbon sr1 done 16:09:10 Venkatrangan Govindarajan proposed a change to integration/test: Reduce the numer of ssh - WIP https://git.opendaylight.org/gerrit/58993 16:10:46 major canadian holiday too 16:10:53 Canada 150! woo 16:11:05 Go Canada! 16:11:32 #info vrpolak asks if we can get SFT back in autorel? 16:12:29 #info should we remove karaf4 from Carbon so we can run SFT in carbon? is it worth the effort to get projects to do this? 16:12:48 Venkatrangan Govindarajan proposed a change to integration/test: Reduce the numer of ssh - WIP https://git.opendaylight.org/gerrit/58993 16:16:41 #action vrpolak to put a patch to enable SFT in autorel for carbon. we can discuss if we want to merge it on the patch 16:17:24 #action zxiiro to check in odlparent on how to flip the switch to not run SFT on k4 in autorel carbon 16:18:47 #info the plan is to just remove the k4 SFT in carbon, then try this in sandbox and if all is well, we will restore it. 16:19:06 #topic distribution 16:20:02 #info all karaf3 is now removed from integration/distribution for nitrogen 16:20:39 #info most nitro csit should not be failing because projects haven't been added to int/dist for k4 yet 16:23:11 #info if projects want to run CSIT on their own local k4 distribution they can hard code the bundle url in to their jjb 16:24:00 #link https://git.opendaylight.org/gerrit/#/c/58987/ <--- patch to hard code bundle-url to test local k4 distribution 16:26:41 #link https://bugs.opendaylight.org/show_bug.cgi?id=8571 <-- Bug causing no configure_cluster.sh in Karaf 4 builds. 16:26:46 #info odlparent bug where k4 distribution doesn't have our extra scripts (e.g. configure cluster, set persistence) 16:29:22 https://docs.google.com/spreadsheets/d/1MYyGLFWN2RzUkJl8XMzXQ-3zWuOrUCQpIS6ORbmf4_U/edit#gid=1274532933 16:37:00 Venkatrangan Govindarajan proposed a change to integration/test: Reduce the numer of ssh - WIP https://git.opendaylight.org/gerrit/58993 16:37:02 #info it seems that featuresBoot has problems and things work better when installing features from the karaf shell 16:37:41 #info the integration group wants to file blocking bugs for these situations where featuresBoot does not work and continue with the assumption that both should work. 16:38:08 #action jamoluhrsen to file blocking bug on netvirt featuresBOot issue 16:39:33 #topic builder 16:40:19 #info we now have a maven release job to help release odlparent 16:41:00 #info ODL sites had cert issues recently, which caused problems that is now temporarily worked around. Andy will give a permanent solution soon 16:41:58 #info RackSpace had some outage over the weekend, but seems to have been resolved. there was some broken battery and broke the jenkins production server 16:43:11 #info vrpolak saw that some longevity jobs failed after 24h (prematurely) so wondering if this could have been the new script to kill stale nodes 16:43:47 zxiiro: are we planning on adding integration/distribution back to autorelease (now that it has removed nonoffset0 projects) 16:44:42 #info we need to reduce longevity job times to something like 23 hours. the zombie killer script is only granular by days 16:44:51 anipbu: yes, was waiting for the go ahead from integration 16:45:00 #action jamoluhrsen to search for longevity timers in int/test to warn projects 16:45:41 LuisGomez: vrpolak: I believe we have go ahead for adding integration/distribution back to autorelease (now that it has removed nonoffset0 projects)? Please let us know. 16:45:49 #link https://jenkins.opendaylight.org/releng/job/controller-csit-3node-periodic-benchmark-only-boron/323/ Controller job hung 16:46:00 #topic packaging 16:46:28 #info nexus repo getting setup for int/pak so we can push artifacts to nexus 16:47:35 +1 >adding integration/distribution back to autorelease 16:48:13 #info major docs update in int/pak that we need to review and merge: https://git.opendaylight.org/gerrit/58670 16:48:20 it's in my to review queue 16:49:51 #info dfarrell07 giving multiple talks at OPFNV Summit on behalf of ODL. -- kudos! 16:51:40 #topic misc 16:52:17 #info some Ericcson folks emailed about S3P. jamoluhrsen will loop in mattw4 16:52:50 Faseela K proposed a change to integration/test: Fix issues with genius script plans https://git.opendaylight.org/gerrit/59039 16:53:16 #info LuisGomez thinks we may have some good intern resources this summer to help us with some of our "good ideas". e.g. vrpolak's banhammer script 16:53:59 #info LuisGomez wants to give a dashboard demo next week 16:57:00 #info mattw4 working on a scale framework with devstack and docker containers. things are stabilizing now and we can consider using these in upstream jobs 16:58:14 #info each devstack container is consuming ~3G of memory 16:59:49 #info mattw4 asking for input/test cases if anyone has it. dfarrell07 points out that afredette has a scale/perf related test doc. 17:00:19 #endmeeting